Soluble and insoluble impurities Snow-pit studies were conducted at three locations during the 1987-1988 field season (the Bird Saddle, the Terra Nova in snow samples Saddle, Windless Bight) and at three additional locations (Fang from , Antarctica Glacier, the Terror Saddle, and the Ross Ice Shelf) during the 1988-1989 field season (see figure 1). We report here only our results from the 1987-1988 pit studies. Seasonal variations in

JULIE M. PALAIS, RAYMOND CHUAN, oxygen-isotope ratios are reasonably well preserved at all three and MARY Jo S1ENcEI sites. The accumulation rates (in terms of centimeters of snow per year) as estimated from the oxygen isotope profiles are as follows: Windless Bight approximately 20-40 centimeters per Glacier Research Group year, Bird Saddle approximately 50-60 centimeters per year, Unioersity of New Ham ps/ore Durham, New Ham ps/ore 03824 Terra Nova Saddle approximately 40-50 centimeters per year. Major anions (sulfate, chlorine, and nitrate) were determined by ion chromatography. In this paper, we report preliminary results of snow-pit and Figure 2 shows our results of oxygen-18, sulfate, chloride, urn-core studies that we are making to examine the types of and nitrate from the Windless Bight pit. The chemical varia- soluble and insoluble particles found in snow samples collected tions seen in this pit are representative of the types of varia- around , the active volcano on Ross Island, Ant- tions seen in all of the pits which we have studied to date. arctica (figure 1). This work is part of a larger study in which The major anion concentrations in the Terra Nova Saddle pit we are also making aerosol measurements aboard an LC-130 are similar to those in the Windless Bight pit. The concentra- Hercules airplane and on the ground in order to determine tions in the Bird Saddle pit, however, are similar for nitrate, whether Mount Erebus has an effect on the chemistry of the but one to two orders of magnitude greater for sulfate and up antarctic troposphere and on regional ice chemistry. The aero- to three orders of magnitude greater for chloride. The source sol measurements are made with a quartz-crystal microbalance of at least some of this sulfate and chloride is believed to be multi-stage cascade impactor to characterize the typical aerosol from sea salt, because abundant sodium chloride crystals were particles found in the volcanic plume of Mount Erebus (both seen in snow samples from this site which were filtered and on the ground at the crater rim and in the air) and in the examined by scanning electron microscope (figure 313). This is ambient troposphere (see Chuan and Palais, Antarctic Journal, probably due to the fact that the Bird Saddle is at a relatively this issue). Several aerosol measurements have also been made low elevation (750 meters) and is very close to areas of open on the ground but away from the direct effects of the volcano. water throughout, at least, the austral summer period. It is interesting to note in the Windless Bight pit profile that there are several large peaks of chloride and sulfate in the years leading up to and including the summer of 1984-1985, which was a period of increased activity at Mount Erebus (Kyle 1986). Some of these peaks appear to show excess chloride (i.e., non-sea-salt chloride) and they are probably volcanic in origin. Snow from these layers is being studied further to look for evidence of volcanic ash or other particles indicative of Erebus activity. Nitrate exhibits maxima in summer at most sites with occasional secondary peaks in winter at some sites. A possible relations with the formation of the ozone hole and the precipitation, in winter, of nitrate from polar stratospheric 30 clouds is being considered to explain these winter nitrate peaks (Legrand and Kirchner 1988). We have been developing new freeze-drying techniques (ly- ophilization) to extract particles from the snow without melt- ing. These particles, along with those collected in the cascade impactor, are examined by scanning electron microscope and energy-dispersive X-ray analysis to determine their morphol- ogy and chemical composition. Studies of particles extracted both by melting and lyophilizing snow-pit and firn-core sam- ples from sites around Ross Island are in progress and com- parisons are being made between the particles found in the snow and those sampled in the atmosphere by cascade im- pactor during the 1987-1988 and 1988-1989 field seasons. Pre- liminary results suggest that the lyophilizing technique preserves a class of amorphous particles composed of low atomic weight (Z) elements such as carbon and oxygen (see figure 3C). Similar particles have been observed in the atmospheric quartz-crystal 65 66 167o 168o I69 microbalance sampling (see Chuan and Palais, Antarctic Jour- Figure 1. Map of Ross Island showing sampling locations from nal, this issue). We also find particles composed of soluble salts 1987-1988 field season (Bird Saddle, Windless Bight, Terra Nova on these filters, such as potassium chloride (figure 3A), sodium Saddle) and 1988-1989 field season (Fang Glacier, Terror Saddle chloride (figure 313), sodium sulfate, and calcium sulfate (figure Ross Ice Shelf). (km denotes kilometer.) 3D).

1989 REVIEW 89 W A snow sample from the Terra Nova Saddle was lyophilized and passed through the cascade impactor so that we could 50 study the mass and size distribution of small particles. The 40 total recovery of particles in the size range from 0.1 to about 20 micrometers came to about 4 x 10 grams per gram of ice. .30 a The particles found on the cascade impactor stages from this sample include complex aggregates (up to about 10 micrometer 20 diameter) of insoluble material composed of silicon, alumi- 10 num, calcium, iron, sulfur, and sodium chloride crystals (ap- proximately 2-micrometer diameter) with small amounts of 01 I I magnesium sulfate, sodium sulfate, calcium sulfate (see figure 100 200 31)), and coalesced sulfuric acid droplets. 14 Examination of particles in the snow pit layers believed to be associated with fallout from Mount Erebus indicate the pres- 12 ence of silicate shards, gold, soluble salts, and oxides of chro- mium and iron, which are all thought to be characteristic of 10 Mount Erebus emissions (see Chuan and Palais, Antarctic Jour- nal, this issue). It is our pleasure to thank the LC-130 and UH-1N crews of VXE-6 for excellent logistical support. This work was funded 0 by National Science Foundation grant DPP 87-04319. 4

2 References

0 Chuan, R., and J.M. Palais. 1989. Do gold, chromium oxide, and 0 100 200 carbon-containing particles provide tracers of Mount Erebus emis- sions? Antarctic Journal of the U.S., 24(5). 8 Kyle, P.R. 1986. Volcanic activity of Mount Erebus, 1984-1986. Ant- arctic Journal of the U.S. 21(5), 7. Legrand, M., and S. Kirchner. 1988. Polar atmospheric circulation and 6 chemistry of recent (1957-1983) south polar precipitation. Geophysical Research Letters, 15, 879-882.

0 z 2

0 (. co co 100 200 N. Ln

a-30 a

-40 Depth (cm) Figure 2. Profiles of isotopic oxygen-18 (180), sulfate (SO42- ), ni- trate (NO3 _), and chloride (Cl-) in Windless Bight snow pit. (1iEq! I denotes microequivalents per liter. cm denotes centimeter.)

90 ANTARCTIC JOURNAL

VM&O F!!

C

Figure 3. Scanning electron micrographs of particles extracted by lyophilization from snow samples collected at sites around Ross Island. A. Potassium chloride (KCI) pyramid—Erebus Crater Rim snow. B. Sodium chloride (NaCI) cubes—Bird Saddle snow. C. Unidentified particle composed of carbon, oxygen, and silicon—Terra Nova Saddle snow. D. Sodium sulfate (Na2SO4) and calcium sulfate (CaSO 4)— sharp, needle-like crystals—Terra Nova Saddle snow.

1989 REVIEw 91